Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation

ORIGINAL ARTICLE KRAS Mutations in Advanced Nonsquamous Non Small- Cell Lung Cancer Patients Treated with First-Line Platinum- Based Chemotherapy Have No Predictive Value Wouter W. Mellema, MD,* Anne-Marie C. Dingemans, MD, PhD, Erik Thunnissen, MD, PhD, Peter J. F. Snijders, MD, PhD, Jules Derks, Daniëlle A. M. Heideman, MD PhD, Robertjan Van Suylen, MD, PhD, and Egbert F. Smit, MD, PhD* Background: Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation is thought to be related with dismal outcome for non smallcell lung cancer (NSCLC) patients. The role of KRAS mutation as a predictor of response to chemotherapy for patients with metastatic NSCLC is poorly understood. Methods: From a retrospective database of two university hospitals, all patients with advanced, nonsquamous NSCLC treated with first-line platinum-containing chemotherapy were selected. Mutation analysis for KRAS was performed and the relation with response to chemotherapy was assessed. Secondary endpoints were its relation with response to progression-free survival (PFS) and overall survival (OS). Results: A total of 161 patients, 94 men and 67 women, were included in this study. Median age was 60 years. The majority of patients (79%) had stage IV disease, of which 60 patients (37%) had a KRAS mutation. Patients with a KRAS mutation had a similar response to treatment as patients with KRAS wild-type (wt) (p = 0.77). Median PFS in KRAS-mutated patients was 4.0 months versus 4.5 months in KRAS wt patients (hazard ratio = 1.3; [95% confidence interval, 0.9 1.8]; p = 0.16). Median OS in patients with KRAS mutation was 7.0 months versus 9.3 months in patients with KRAS wt (hazard ratio = 1.2; [95% confidence interval, 0.9 1.7]; p = 0.25). Type of KRAS mutation had no influence on response or outcome. Conclusion: On the basis of our multicenter data presented here, we conclude that KRAS mutation is not predictive for worse response to chemotherapy, PFS, and OS in advanced NSCLC patients treated with platinum-based chemotherapy in first-line setting. Key Words: Non small-cell lung cancer, Kirsten rat sarcoma viral oncogene homolog mutation, Predictive biomarker, Prognostic biomarker. (J Thorac Oncol. 2013;8: 1190-1195) *Department of Pulmonary Diseases, VU University Medical Center, Amsterdam, The Netherlands; Department of Pulmonary Diseases and GROW-School for Oncology and Developmental Biology, Maastricht University Medical Center, Maastricht, The Netherlands; Department of Pathology, VU University Medical Center, Amsterdam, The Netherlands; and Department of Pathology, Maastricht University Medical Center, Maastricht, The Netherlands. Disclosure: The authors declare no conflict of interest. Address for correspondence: Egbert F. Smit, MD, PhD. Department of Pulmonary Diseases, VU University Medical Center, PO Box 7057, 1007 MB, Amsterdam, The Netherlands. E-mail: ef.smit@vumc.nl Copyright 2013 by the International Association for the Study of Lung Cancer ISSN: 1556-0864/13/0809-1190 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation is the most common driver mutation in non small-cell lung cancer (NSCLC) and is present in approximately 30% of all NSCLC patients, mainly in patients with adenocarcinoma (AC) histology. Other mutations in rat sarcoma viral oncogene homolog (RAS) gene family include mutations in Harvey rat sarcoma viral oncogene homolog (HRAS) and neuroblastoma rat sarcoma viral oncogene homolg (NRAS) but these are observed in less than 10% of the RAS mutations in NSCLC. RAS proteins circulate signals from growth receptors on the cell surface to intracellular effector pathways responsible for cell growth and proliferation. Mutations most often occur in codons 12, 13, or 61 of the RAS gene (located on chromosome 12) and result in an irreversible continuous activation of RAS protein. 1 3 Because of the critical role of RAS protein in cell proliferation, NSCLC patients with a KRAS mutation are believed to have a worse prognosis as compared with patients with a KRAS wild type (wt). The relation between KRAS mutational status and survival was first reported in 1990. 4 Since then, a variety of studies have investigated the influence of KRAS mutational status on survival, however, even today, conflicting results are reported. It is believed that patients with a KRAS mutation do not respond to chemotherapy treatment, although at this moment, chemotherapy is the only treatment option for these patients. 5 In contrast, successful targeted agents are developed for patients with an EGFR mutation and EML4-ALK translocation. 6,7 Few studies have investigated the effect of KRAS mutation on response to first-line chemotherapy in patients with advanced NSCLC. 8 10 These studies conclude that patients with a KRAS mutation do not have worse response to chemotherapy treatment. Though, because of the small sample size and differences in type of chemotherapy used in these studies, this finding remains open to debate. The aim of our retrospective study is to evaluate the association of KRAS mutational status with response to chemotherapy, progression-free survival (PFS), and overall survival (OS) in patients with advanced NSCLC treated with platinum-based chemotherapy as first-line treatment. PATIENTS AND METHODS Study Patients We retrospectively selected all consecutive nonsquamous (p63-negative) NSCLC patients by searching pathology reports 1190 Journal of Thoracic Oncology Volume 8, Number 9, September 2013

Journal of Thoracic Oncology Volume 8, Number 9, September 2013 KRAS Mutations in Advanced Nonsquamous, NSCLC in two university hospitals from 2004 to January 2011. Patients were eligible for the study on the basis of the following criteria: proven incurable stage IIIb or IV (tumor, node, metastasis [TNM] 6th edition) NSCLC, treatment with first-line platinumcontaining chemotherapy, availability of tumor material for KRAS mutation analysis, at least one target lesion according to Response Evaluation Criteria in Solid Tumours 1.0, and evaluable computed tomography before and after treatment. Palliative radiotherapy during chemotherapy treatment was allowed. The following data were retrieved from the medical records: age, sex, smoking history, World Health Organisation performance status (PS), histology (including p63 staining results), stage of disease (Union for International Cancer Control [UICC] 6th edition), chemotherapy treatment, number of courses, response to treatment according to the Response Evaluation Criteria in Solid Tumours 1.0, date of start of treatment, date of progression, and date of death or date of last contact. Ethics Statement This study was approved by the ethics committee of the VU University Medical Center Amsterdam; according to our local rules, informed consent was not needed for this retrospective study. Immunohistochemistry To distinguish between AC and squamous cell carcinoma and thyroid transcription factor 1 (TTF1) and p63, immunohistochemistry staining was performed on all patients, according to a previously described protocol. 11 For TTF1 clones 8G7G3/1 were used and for p63 clones 4A4 were used. A single observer scored the stained slides for intensity (0 = negative, 1 = weak, 2 = moderate, and 3 = strongly positive) and percentage of positive tumor cells with 10% increments. For each immunohistochemical stain, a total score was obtained by multiplying intensity and percentage of positive tumor cells. A score more than 240 was considered positive for p63. This threshold was chosen because it clearly delineates squamous cell carcinoma from AC differentiation. 12 For TTF1 a positive threshold of more than 30 was used, as even minor staining is associated with AC differentiation. Mutation Analyses Tumor tissue was manually macrodissected from serial sections guided by a hematoxylin-eosin stained tissue section, on which the tumor area was marked by a pathologist. DNA was extracted and subjected to high-resolution melting and sequencing analysis for KRAS exon 2 and 3, according to routine protocol. 13,14 Statistical Analysis The primary objective was to evaluate the difference in probability of response between patients with KRAS wt and KRAS mutation. Secondary objectives were PFS, defined as time from start of treatment till objective disease progression or death, and OS defined as the time from start of treatment till death. The relationship between KRAS mutational status and patient characteristics, type of chemotherapy treatment, or response was calculated using Pearson s χ 2 test. Kaplan Meier curve was used to estimate the distribution of survival according to KRAS mutational status. Log-rank test was used to calculate difference in survival among the subgroups. To estimate the hazard ratio (HR), Cox regression analysis was used. In addition, analysis was performed for different types of KRAS mutation. For the latter analysis, types of KRAS mutations that had a frequency of less than 5 were clustered in one group. RESULTS From the retrospective databases, 161 patients were eligible for the study. The patient characteristics are listed in Table 1. The median age was 60 years (range, 34 83 years), 17 patients (10.3%) had a PS of more than 1. The majority of patients had stage IV disease (79%). A total of 115 patients (71.4%) had an AC, 46 patients (29.6%) were diagnosed with NSCLC not otherwise specified favoring AC. Treatment According to the inclusion criteria, all patients received a platinum combination. Eighty-nine patients were treated with cisplatin, 69 patients were treated with carboplatin, and three patients received both regimens. Platinum was most frequently combined with gemcitabine (51.6%). Other partners were pemetrexed (31.1%), docetaxel (16.8%), and vinorelbine (0.6%). The median amount of chemotherapy courses was four. In total, there was one patient (0.6%) with a complete remission, 31 patients (19.3%) had a partial response (PR), 77 patients (47.8%) had a stable disease (SD), and 52 patients (30.7%) had progressive disease (PD). KRAS Mutation KRAS mutations were present in 60 patients (37.3%), of which the majority of patients had stage IV disease (85.0%). There was no significant relationship among patient characteristics as listed in Table 1 and KRAS mutational status. Also, there were no differences in treatment among patients with or without a KRAS mutation (p = 0.90) or in amount of courses administered (p = 0.31). For KRAS wt patients, platinum doublet chemotherapy resulted in 1 complete remission (1.0%), 21 PRs (20.8%), 48 SDs (47.5%), and 31 PDs (30.7%). Responses in the KRAS-mutated group included 10 PRs (16.7%), 29 SDs (48.3%), and 21 PDs (35.0%; p = 0.77) (Fig. 1). The different types of mutation are presented in Figure 2. The most frequent mutations observed were: p.g12c (41.7%), p.g12v (16.7%), p.g12d (11.7%), and p.g13c (10.0%). There were no differences between type of mutation and response to therapy (p = 0.98). PFS and OS The median follow-up was 48 months. Date of death was assessed on October 1, 2011. Median PFS in KRASmutated patients was 4.0 months (95% CI, 2.8 5.2 months) versus 4.7 months (95% CI, 3.2 6.1 months) in KRAS wt patients (HR = 1.3; [95% CI, 0.9 1.8]; p = 0.12) (Fig. 3A). Median OS in patients with KRAS mutation was 7.0 months (95% CI, 3.9 10.2) versus 9.3 months (95% CI, 6.6 11.9) in patients with KRAS wt (HR = 1.2; [95% CI, 0.9 1.7]; p = 0.25) (Fig. 3B). Classic prognostic factors, such as stage Copyright 2013 by the International Association for the Study of Lung Cancer 1191

Mellema et al. Journal of Thoracic Oncology Volume 8, Number 9, September 2013 TABLE 1. Patient Characteristics KRAS Wild-Type KRAS Mutation n % n % Sign (p < 0.05) Median, age (range, yr) 61 (34 80) 58 (35 83) 0.61 Sex 0.33 M 56 55.4 38 63.3 F 45 44.6 22 36.7 Smoking history 0.09 Never 12 13.5 2 3.5 Former 41 46.1 25 43.9 Current 36 40.4 30 52.6 WHO PS 0.61 0 43 45.7 23 40.4 1 40 42.6 28 49.1 2 9 9.6 6 10.5 3 2 2.1 0 0.0 Histology 0.68 Adenocarcinoma 71 70.3 44 73.3 NSCLC-NOS 30 29.7 16 26.7 Stage 0.18 IIIb 24 23.8 9 15.0 IV 77 76.2 51 85.0 Platinum regimen 0.91 Carboplatin 2 2.0 1 1.7 Cisplatin 42 41.6 27 45.0 Both 57 56.4 32 53.3 Chemotherapy regimen 0.90 Gemcitabine 52 51.5 31 51.7 Pemetrexed 31 30.7 19 31.7 Docetaxel 17 16.8% 10 16.7 Vinorelbine 1 1.0 0 0.0 WHO PS, World Health Organisation performance status; NSCLC-NOS, non small-cell lung cancer not otherwise specified. FIGURE 1. Histogram of response (in %) in patients with KRAS wt or KRAS mut. CR, complete remission; KRAS, Kirsten rat sarcoma viral oncogene homolog; mut, mutation; PD, progressive disease; PR, partial response; SD, stable disease; wt, wild type. (HR = 1.6; [95% CI, 1.0 2.5]), histology (AC versus large cell; HR = 1.6; [95% CI, 1.1 2.3]), and PS (HR = 1.5; [95% CI, 1.2 2.0]) were found to be prognostic factors. The 1-year survival was 36.7 and 46.1% in KRAS-mutated and KRAS-wt patients, respectively. The PFS and OS were comparable in the different types of KRAS mutation (p = 0.90 and p = 0.99, respectively). DISCUSSION In this retrospective study of a consecutive cohort of patients, we observed no differences in response to firstline platinum-based chemotherapy treatment in advanced NSCLC patients with or without a KRAS mutation. Although median OS and the 1-year survival rate was worse in patients 1192 Copyright 2013 by the International Association for the Study of Lung Cancer

Journal of Thoracic Oncology Volume 8, Number 9, September 2013 KRAS Mutations in Advanced Nonsquamous, NSCLC FIGURE 2. Histogram of type of KRAS mutation. KRAS, Kirsten rat sarcoma viral oncogene homolog. harboring a KRAS mutation, this was not statistically significantly different. The aggressive behavior of KRAS mutation in NSCLC patients was first suggested by Slebos et al. 4 in 1990, who reported worse survival of KRAS-mutated NSCLC patients. The hypothesis is provided by the function and downstream effectors of the RAS protein family, including the mitogenactivated protein kinase (MAPK) pathway. An activating KRAS mutation could, therefore, result in continuous stimulation of tumor proliferation, resulting in early progression and poor survival. In accordance with other studies, we found that response and PFS were not significantly worse in advanced NSCLC patients with a KRAS mutation. (Table 2). In a prospective trial, 62 patients with inoperable stage III or IV NSCLC were treated with mesna, ifosfamide, carboplatin, and etoposide. Sixteen patients (25.8%) had a KRAS mutation. In three of 16 patients (19%) with a KRAS mutation there was response to treatment, compared with a response rate of 26% in KRAS-wt patients. This difference in response was not significant (p = 0.49). The PFS and OS were not significantly different in patients with KRAS wt and KRAS mutation. 8 Another study evaluated mutational status and response to treatment using data of the TRIBUTE trial. This was a randomized phase III study in advanced NSCLC, comparing first-line chemotherapy with first-line chemotherapy with concurrent erlotinib. Of 274 patients, 264 had tumor material available for KRASmutation analysis. In 55 patients (21%) a KRAS mutation was present. No differences in response were seen between KRASmutated patients treated in the erlotinib-containing arm and the chemotherapy-only arm (8% versus 23%, respectively; p = 0.16). Also, no differences in PFS and OS was observed in patients either with, or without a KRAS mutation. 9 Recently, a study evaluated clinical outcome in advanced NSCLC patients receiving first-line chemotherapy according to EGFR and KRAS mutational status. In this study, 162 patients were treated with first-line chemotherapy. Thirty of 133 patients (22.6%) had a KRAS mutation. No difference in response to chemotherapy was found between patients with KRAS mutation or KRAS wt (25.0% versus 26.5%, respectively; FIGURE 3. A, PFS in months. Continuous line KRAS wt (median PFS 4.7 months), dashed line KRAS mut (median PFS 4.0 months). HR = 1.3 (95% CI, 0.9 1.8). B, OS in months. Continuous line KRAS wt (median OS 9.3 months), dashed line KRAS mut (median OS 7.0 months). HR = 1.2 (95% CI, 0.9 1.7). CI, confidence interval; HR, hazard ratio; KRAS, Kirsten rat sarcoma viral oncogene homolog; mut, mutation; os, overall survival; PFS, progression-free survival; wt, wild type. Copyright 2013 by the International Association for the Study of Lung Cancer 1193

Mellema et al. Journal of Thoracic Oncology Volume 8, Number 9, September 2013 TABLE 2 Predictive Value of KRAS Mutational Status Response Study Histology Type of Chemotherapy No. of Patients a Mut (%) KRAS KRAS Mut (%) KRAS Wt (%) Predictive Rodenhuis et al. 1997 8 Adenocarcinoma Mesna, ifosfamide, carboplatin, 62 26 19 26 No and etoposide Eberhard et al. 2005 9 All Carbo/paclitaxel ± erlotinib 264 21 23 26 No Kalikaki et al. 2010 10 All Several 133 23 25 27 No a Number of patients eligible for KRAS mutation analysis. Wt, wild type; Mut, mutation; KRAS, Kirsten rat sarcoma viral oncogene homolog. p = 0.87). In patients treated with platinum-based chemotherapy (96 patients, of which 18 patients were with a KRAS mutation) there were also no differences in response (29.2% versus 30.2%, respectively; p = 0.95). The PFS and OS was comparable in patients with or without a KRAS mutation. 10 A variety of studies have investigated the prognostic role of KRAS mutational status. Unfortunately, these studies were equivocal because of differences in patient selection, stage of disease, histology, and type of treatment. In the early 1990s, KRAS mutation was reported to be a poor prognostic factor in early-stage NSCLC patients who underwent surgery. 4,15,16 However, in subgroup analysis of prospective studies in NSCLC, KRAS mutational status was not found to be prognostic. 17,18 In a meta-analysis on the prognostic role of KRAS mutational status in NSCLC, reviewing 53 studies and 5216 patients, KRAS mutational status was found to be a poor prognostic factor. Patients with a KRAS mutation had a worse survival with an HR of 1.40 (95% CI, 1.18 1.65). 19 Unfortunately, a multivariate analysis with prognostic factors such as PS and stage of disease was not performed. A post hoc power analysis for response rate and 1-year survival rate (SAS 9.2; β = 0.9; α = 0.05; weight is 3 for KRASmutated patients and 5 for KRAS wt) demonstrated that with our sample size, a reduction of 20% in 1-year survival could be detected with a power of 0.80. Prospective studies should be conducted to answer the question whether KRAS mutational status has prognostic or predictive value. However, based on the studies discussed above and our findings, KRAS mutational status does not seem to have a clinically relevant impact on PFS and OS. Therefore, the value of KRAS mutational status in NSCLC as predictor of poor outcome has to be reviewed. A very interesting hypothesis is that different types of KRAS mutation are associated with different responses to chemotherapy regimens. Garassino et al. 20 described this in a study on KRAS-mutated NSCLC cell lines. Differences were found in three types of KRAS mutation: p.g12c, p.g12v, and p.g12d. In comparison with the WT clones, p.g12c mutation was associated with a reduced response to cisplatin, but increased sensitivity to taxol and pemetrexed, whereas p.g12v mutation showed a strong sensitivity to cisplatin, less sensitivity to pemetrexed. Cell lines harboring a p.g12d mutation showed resistance to taxol, but sensitivity to sorafenib. Another recently published study reported poor survival in patients with KRAS p.g12c and p.g12v, treated with molecular targeted therapy compared with other types of KRAS mutation. 21 This study used data of the Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination (BATTLE) trial, in which a total of 255 patients were randomized to erlotinib, vandetanib, bexarotene, and erlotinib or sorafenib. 22 A total of 43 patients had a KRAS mutation; of these, 24 patients had a p.g12c or p.g12v mutation. KRAS mutational status was not associated with OS or PFS (p = 0.09). When KRAS mutations were clustered in p.g12c + p.g12v and others, there was a significant difference in median PFS when compared with KRAS wt (1.8, 3.4, and 2.0 months, respectively; p = 0.046). In our patient population, we could not confirm these data, but the sample size was probably too small to find relevant differences between the types of KRAS mutation. We encourage further study on type of KRAS mutation and its relation to response to chemotherapy and prognosis in a larger cohort of patients. CONCLUSION On the basis of our multicenter data presented here, we conclude that KRAS mutational status is not likely to be predictive for worse response or PFS in advanced nonsquamous NSCLC patients treated with platinum-based chemotherapy in first-line setting. REFERENCES 1. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57 70. 2. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011;144:646 674. 3. Pylayeva-Gupta Y, Grabocka E, Bar-Sagi D. RAS oncogenes: weaving a tumorigenic web. Nat Rev Cancer 2011;11:761 774. 4. Slebos RJ, Kibbelaar RE, Dalesio O, et al. K-ras oncogene activation as a prognostic marker in adenocarcinoma of the lung. N Engl J Med 1990;323:561 565. 5. Adjei AA. K-ras as a target for lung cancer therapy. J Thorac Oncol 2008;3(6 Suppl 2):S160 S163. 6. Kwak EL, Bang YJ, Camidge DR, et al. Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer. N Engl J Med 2010;363:1693 1703. 7. Lynch TJ, Bell DW, Sordella R, et al. Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib. N Engl J Med 2004;350:2129 2139. 8. Rodenhuis S, Boerrigter L, Top B, et al. Mutational activation of the K-ras oncogene and the effect of chemotherapy in advanced adenocarcinoma of the lung: a prospective study. J Clin Oncol 1997;15:285 291. 9. Eberhard DA, Johnson BE, Amler LC, et al. Mutations in the epidermal growth factor receptor and in KRAS are predictive and prognostic indicators in patients with non-small-cell lung cancer treated with chemotherapy alone and in combination with erlotinib. J Clin Oncol 2005;23:5900 5909. 10. Kalikaki A, Koutsopoulos A, Hatzidaki D, et al. Clinical outcome of patients with non-small cell lung cancer receiving front-line 1194 Copyright 2013 by the International Association for the Study of Lung Cancer

Journal of Thoracic Oncology Volume 8, Number 9, September 2013 KRAS Mutations in Advanced Nonsquamous, NSCLC chemotherapy according to EGFR and K-RAS mutation status. Lung Cancer 2010;69:110 115. 11. Galleges Ruiz MI, Floor K, Steinberg SM, et al. Combined assessment of EGFR pathway-related molecular markers and prognosis of NSCLC patients. Br J Cancer 2009;100:145 152. 12. Rossi G, Papotti M, Barbareschi M, Graziano P, Pelosi G. Morphology and a limited number of immunohistochemical markers may efficiently subtype non-small-cell lung cancer. J Clin Oncol 2009;27:e141 2; author reply e143. 13. Kramer D, Thunnissen FB, Gallegos-Ruiz MI, et al. A fast, sensitive and accurate high resolution melting (HRM) technology-based assay to screen for common K-ras mutations. Cell Oncol 2009;31:161 167. 14. Heideman DA, Lurkin I, Doeleman M, et al. KRAS and BRAF mutation analysis in routine molecular diagnostics: comparison of three testing methods on formalin-fixed, paraffin-embedded tumor-derived DNA. J Mol Diagn 2012;14:247 255. 15. Rosell R, Li S, Skacel Z, et al. Prognostic impact of mutated K-ras gene in surgically resected non-small cell lung cancer patients. Oncogene 1993;8:2407 2412. 16. Fukuyama Y, Mitsudomi T, Sugio K, Ishida T, Akazawa K, Sugimachi K. K-ras and p53 mutations are an independent unfavourable pro gnostic indicator in patients with non-small-cell lung cancer. Br J Cancer 1997;75:1125 1130. 17. Schiller JH, Adak S, Feins RH, et al. Lack of prognostic significance of p53 and K-ras mutations in primary resected non-small-cell lung cancer on E4592: a Laboratory Ancillary Study on an Eastern Cooperative Oncology Group Prospective Randomized Trial of postoperative adjuvant therapy. J Clin Oncol 2001;19:448 457. 18. Broermann P, Junker K, Brandt BH, et al. Trimodality treatment in Stage III nonsmall cell lung carcinoma: prognostic impact of K-ras mutations after neoadjuvant therapy. Cancer 2002;94:2055 2062. 19. Mascaux C, Iannino N, Martin B, et al. The role of RAS oncogene in survival of patients with lung cancer: a systematic review of the literature with meta-analysis. Br J Cancer 2005;92:131 139. 20. Garassino MC, Marabese M, Rusconi P, et al. Different types of K-Ras mutations could affect drug sensitivity and tumour behaviour in nonsmall-cell lung cancer. Ann Oncol 2011;22:235 237. 21. Ihle NT, Byers LA, Kim ES, et al. Effect of KRAS oncogene substitutions on protein behavior: implications for signaling and clinical outcome. J Natl Cancer Inst 2012;104:228 239.m 22. Kim E, Herbst R, Wistuba I, et al. The BATTLE Trial: personalizing therapy for lung cancer cancer discovery 2011 June 2011;1:44 53. Copyright 2013 by the International Association for the Study of Lung Cancer 1195